The overall aim of my dissertation was to investigate the effect of individual and collective characteristics on team performance. In this dissertation, the scientific question was first stated, and the notions related to our question including collective problem solving, agent-based modeling, and NK modeling were reviewed.
Second, it was determined how to develop an agent-based model in which networked agents collectively search for solutions to simple and complex problems in which the solution space is represented by an NK landscape. Teams of agents share information about newly found solutions while exploring the NK landscape. Running simulations on the model, the impact of team members’ behavioral strategies in addition to team size and diversity on the team performance both for situations in which a team faces simple problems and complex problems were studied. The behavioral strategies that were studied include normal, hardworking, and risk-taking. While normal agents only explore one solution per unit time, hard-working agents explore more than one solution per unit time, and risk-taking agents explore one solution, but injecting randomness into the solution makes that solution substantially different from their current solution.
It was shown that when teams deal with complex problems, both risk-taking and hardworking strategies improve the teams’ final scores with moderate risk-taking has a substantial positive impact on teams’ performance. Hardworking strategy has a positive effect on how quickly a team can reach its final solution while risk-taking strategy has a negative effect on it. On the other hand, when a team faces simple problems, risk-taking strategy has a negative impact on team performance, but hard work reduces the amount of time it takes to solve a problem. It was also shown that a larger team can solve problems more effectively; however, some parts of this positive effect might occur because, in our model, the larger teams are more diverse. Therefore, the effect of diversity on team performance was studied and an increasing team diversity was shown to lead to an improvement in the ultimate scores of the teams, and that more diverse teams took less time to reach their final solution.
Third, the effect of superiority bias and communication noise on collective problem solving was studied. Hardworking and risk-taking behavior that were addressed in the second chapter are intentional characteristics of team members. In the third chapter, the effect of two unintentional characteristics on team performance were studied: (1) superiority bias and (2) communication noise.
When agents have a superiority bias, they prefer their own solution to those offered by others. In other words, superiority bias is a reluctance to adopt others’ marginally better solutions. It was shown that both superiority bias and communication noise can lead teams to find higher-quality solutions to complex problems but at the cost of longer search times. It was also demonstrated that a moderate level of communication noise can be optimal for both minimizing the time and resources needed to solve simple problems in comparison to both high and low levels of noise. Increasing the connectedness of teams’ communication network, decreasing the time and resource consumption in addition to the final solutions’ score. When a team faces simple problems, teams with different levels of noise and bias can reach the most optimum solution. However, as a team expresses more superiority bias, it requires more time and resources to reach the final solution. In contrast, a moderate level of communication noise can cause a decrease in resources and time a team needs to converge to its final solution. It was also found that risk-taking teams showed more robustness in terms of resource consumption when teams dealt with noise and bias. This study contributes to the greater body of knowledge regarding collective problem-solving in teams.
Lastly, in Chapter Four, we concluded that the phenomena that maintain solution diversity on a team prolong the team’s convergence to finding a solution. In turn, this causes teams to spend more time exploring the NK landscape. This process results in a better solution at the expense of time and resources. We explained how these findings align with empirical studies and how they can be beneficial to team managers. We also explained the room for future studies and described how there are opportunities for more discoveries along this line of research.